Spatial Models as Powerful Tools for Climate Change Ecology
Phillips, Austin Joseph
MetadataShow full item record
Many scientists suggest that the Earth has entered a new geologic epoch, the Anthropocene, defined by human influence on nearly every physical and biological system. Climate change is perhaps the most widespread human impact. One ecological effect of climate change is the movement of species' habitats due to changes in temperature, precipitation, and other environmental variables. Populations may either track their habitats using dispersal, acclimate to new conditions with phenotypic plasticity, or adapt in place by rapid evolution. Though paleontological records give some clues about past responses to abrupt environmental change, the rapid pace of current climate trends leaves the fate of many species somewhat unknown. Ecologists must, therefore, rely on predictive models to anticipate the dynamics of populations adjusting to their moving habitats. In this work, I focus on a number of spatial models that describe populations responding to their moving habitats. In the first two studies, I use integrodifference equations to describe populations that grow and disperse in distinct stages. Researchers have already used integrodifference equations to model populations tracking their habitats. My goal is to extend their work by (a) adding spatial realism and (b) comparing the long-term and short-term population dynamics that can be observed. In a third study, I use an optimization model based on reserve design to explore optimal management strategies for a population tracking its habitat. The model offers a prescriptive plan for allocating conservation resources under a limited budget. In the first study, I model a population in two-dimensional space. I find that the shape of the population's moving habitat has a strong influence on the population's fate. The pattern of dispersal determines whether the population does best in a long habitat, a wide habitat, or a square habitat. Moreover, the study shows that simplified spatial models can greatly overestimate persistence compared to spatially realistic ones. In the second study, I focus on the short-term dynamics of a population tracking its habitat. The model shows that a mismatch can occur between short-term and long-term population trends, which creates a confounding effect for prediction and management. I focus on understanding when and why such a mismatch occurs, depending on the population's biology, environment, and spatial distribution. Finally, the third study highlights that management of vulnerable species during climate change must be dynamic. I focus on an at-risk population that grows, disperses, and acclimates as temperature increases over time. The model identifies, among a collection of habitat sites, the sequence of management actions that maximizes a conservation metric related to total population size. While many models incorporate dispersal into management decision analysis, this study makes an important advancement by including acclimation in an explicit way. My approach with these three studies is to develop models that (a) advance the mechanistic understanding of population dynamics during climate change, (b) are easy to parameterize with data from a wide variety of species, and (c) make testable predictions. In doing so, I hope to inspire further interdisciplinary work between theoretical and applied ecologists as we confront the challenges of climate change.